About 30% of uveal melanoma patients die from metastatic disease within 5 years. The primary target of metastasis is the liver, and the median survival after a diagnosis is 4 to 7 months. There is current no standard treatment, and better therapies are needed to improve patient survival. It will be necessary to rigorously test new treatments in clinically relevant animal models that mimic the human disease. Ideally these primary and metastatic uveal melanoma models should allow tumor size to be measured over time. This would permit treatment to begin at a standard tumor size and allow tumor growth to be monitored during therapy. In current models of orthotopic human uveal melanoma, it is not possible to continuously monitor tumor growth, because immunocompromised animals cannot be readily moved to standard imaging equipment. The development of a hand-held, high-resolution, high-frequency ultrasound (HF-US) system has made it feasible to attempt to image tumor growth in the eye and liver of nude rats. In this study, the following overall hypothesis will be tested: Using HF-US and a new model of primary choroidal melanoma, [unreadable] differences in tumor progression can be correlated to the presence of periodic acid-Schiff base (PAS)-positive loops in the primary tumor. It will be addressed in 2 specific aims (SA) which hypothesize that: 1) HF-US imaging can be repeatedly used to monitor primary tumor growth and the extent of metastatic disease in nude rat models of primary and metastatic human choroidal melanoma and 2) nude rat survival, levels of plasma tumor markers, and the number of metastatic lesions will be related to the presence of PAS-positive loops, networks, and cross-linked parallel channels in the primary human choroidal melanoma at the time of enucleation. In SA1 HF-US will be innovatively used to noninvasively image primary tumor growth and hepatic metastatic disease in nude rat models of primary and metastatic human uveal melanoma. In SA2 the HF-US system will be used to image primary melanomas in nude rats that reliably reproduce the PAS-positive vascular patterns associated with patient survival. After the tumor-bearing eyes have been enucleated, biweekly HF-US images of the liver will be obtained to track metastatic progression. If rats with primary tumors that show PAS-positive loops have a higher liver metastatic burden as determined by HF-US, this model could be used to study the importance of these patterns in tumor growth and in treatment response. In summary, this project should reveal the applicability of powerful new techniques to study the growth and treatment response of primary and metastatic uveal melanoma in models that mimic the human disease. [unreadable] [unreadable] [unreadable]